Objective: The standard measure of pancreatic enzyme replacement therapy (PERT) efficacy in treating exocrine pancreatic insufficiency (EPI) is the coefficient of fat absorption (CFA). CFA measurement involves 72-hour stool collection, which presents a logistical challenge because, although the test may be performed on an outpatient basis in clinical practice, hospitalization is needed if assurance of complete collection and 100% compliance is required, for example, in controlled situations such as clinical trials. Our aim was to investigate sparse stool sample collection as an alternative to complete 72-hour collection for measurement of stool fat in subjects with EPI.
Subjects and Methods: Prospective data analysis from a previously published, double-blind, randomized, placebo-controlled, 2-period crossover trial in subjects ages 7 to 11 years with EPI caused by cystic fibrosis. Percentage fat (PF) data from sparse stool samples were compared with 72-hour CFA values as a dichotomous variable (<80%, ≥80%), with evaluation of sensitivity, specificity, and positive predictive value. Area under the curve values were obtained from receiver operating characteristic plots of sensitivity versus 1–specificity.
Results: Twelve subjects provided samples for this analysis. Multiple-sample PF values ≤30% were greatly predictive for CFA values ≥80%, as shown by positive predictive value, sensitivity, and specificity values ≥0.89, with high accuracy (AUCs ≥0.93).
Conclusions: Sparse stool sampling for PF analysis appears to be a valid, practical alternative to 72-hour CFA determination and has potential as a screening tool in clinical practice to identify both suboptimal dosing in subjects with EPI receiving PERT and substantial fat malabsorption in subjects not receiving PERT.